sales-operations-planning
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted user-supplied text within its planning logic, creating a surface for indirect prompt injection.
- Ingestion points: Data enters the agent context through various methods in
SKILL.md, includingDemandReview.add_adjustment(reason field),DemandReview.add_assumption(assumption_text field),ExecutiveSopMeeting.add_decision(decision_text field), andExecutiveSopMeeting.add_action_item(description field). - Boundary markers: There are no explicit boundary markers or instructions to the agent to disregard potential instructions embedded within these data fields.
- Capability inventory: The skill demonstrates capabilities for complex data manipulation and visualization using Python. While no network or sensitive file system operations are present in the code, the processed data is used to generate summaries and reports that could influence subsequent agent actions.
- Sanitization: The provided Python classes do not implement sanitization or validation for the free-text fields before they are integrated into the planning objects.
Audit Metadata